> With some languages > requiring restructuring of the phraseology, what seems to be the first > fragment and the second fragment may no longer be ordered in the same > manner in a target language.
That is theoretical, right? For a specific message, and a specific set of languages, I'm fairly optimistic that you can split the message so that you can translate the fragments individually, and then paste them together again so that it still "works" in all languages. > Hence the code that pastes the English > fragments back together may be wrong for a target language. Perhaps. However, if you change the pasting, you should be able to find a solution that works for all languages. > Further, > multiple target languages could have conflicting reorderings, such that > it is nearly impossible to write a single piece of code to paste the > fragments back together. Example? Also, what makes you believe that your approach will solve this problem? IIUC, it should be fairly straight-forward to map your syntax to gettext. It seems that you propose to have multiple placeholders in a skeleton sentence, each placeholder potentially substituting different depending on the value of a parameter. I think this can be mapped fairly easily onto gettext: make one message for each of your placeholders, and another message for the skeleton. IOW, they both have the same expressiveness. As another note: I don't think your ord() function can work. In some languages, the ordinal words depend on the gender of the noun. E.g. Russian: "the first man" - "Первый мужчина" / "Первый человек" "the first woman" - "Первая женщина" Here, "ый" indicates male, "ая". Things get further complicated by case, e.g. German (likewise for Russian) "the first man" - "der erste Mann" "for the first man" - "für den ersten Mann" Regards, Martin _______________________________________________ Python-3000 mailing list Python-3000@python.org http://mail.python.org/mailman/listinfo/python-3000 Unsubscribe: http://mail.python.org/mailman/options/python-3000/archive%40mail-archive.com